1
CHAPTER 1
INTRODUCTION
This section basically will view on the introduction of the thesis. A brief explanation on the background of the thesis and followed by a problem statements defined based
on the theme of the research and an objectives focus of this thesis underlined by the limitation. The limitations are underlined on the designed scope for this research. A
structure on the organization of this thesis research will be given briefly.
1.1 Overview of Research
Neural network provide an application that can be applied in a broad range. It is a powerful new technique for solving problems in many different disciplines. This
theme of research basically will focus onto two different things. At the end of this research will give a correlation in both themes which are difference views; those
areas are multivariate regression and related to neural network existing application to express a neural regression.
2 Linear and nonlinear regression methods are most likely used to modeling the
mathematical model of regression computation. Based on several methods come up from the mathematicians, it can be presented by using this capabilities of
mathematical model in the computation. This because regression refer as the problem to model a continuous dependent variables as a continuous function and can be
possible to independent variables also. Therefore, classis model used to present linear and nonlinear of regression problem.
In order to relate with this research themes, neural network is a method used to develop and design for a regression problem. Hence, the main purpose of this
research is to demonstrate the optimum use of artificial neural network ANN as a soft computation tool for determining the multivariable input and output interrelation
in order determine the function for regression. As with any modeling tool, to build a model that is effective need a lot of
preparation. This preparation involves specifying the model, determining the multivariable data involved and justify the model with a sample case of an extracted
data to be test. Uys, 2010. The concern in this context of research is often many techniques and methods that
are used in these preparation to compute multivariable data. Therefore in this multivariate regression analysis using an artificial neural network, several models are
proposed previous study will be view as the literature for methods used.
1.2 Problem Background
